Superstition is acting as though a behavior will have some sort of causal influence on the world, even though it can be demonstrated it has no such effect. Superstitious behavior is learned when there is an apparent causal link between some action and a desired outcome; in sports these superstitions are cherished as part of overarching traditions, employed in a kind of tongue-in-cheek manner for the purpose of group solidarity. There are even times when a superstitions behavior is correlated to a desirable outcome, in the sense that it acts as a psychological primer for success. Some superstitions behaviors do have observable influence, save that they don’t have the influence attributed to them.
The behavioral psychologist B.F. Skinner demonstrated that it is possible to teach pigeons superstition, by simply exploiting how behaviors can be conditioned. Indeed, it is apparently more difficult to get pigeons to not be superstitious. When a desired outcome occurs, the association is immediately formed that whatever behavior was being done at that time must be the cause, and so the pigeon will continue to perform that behavior as long as it appears to be rewarded for it.
The same kind of conditioning takes place in humans as well, through the accidental association of certain behaviors and desirable outcomes. In particular, there tends to form a causal association between the use of certain words, or appeals to certain values, and the expected influence this is intended to have on someone’s behavior. Given influencing someone else’s behavior through our use of signs is always a relatively open ended affair, it is possible to note when someone is under the impression that another has not responded appropriately in the way desired when they choose to amplify the power of their signal.
This is related to the concept of token-phenomena distribution, in which a powerful word has its meaning extended to encompass a wider range of phenomena; at first the connoted seriousness provides a short term gain to using the word that way and pushing whatever sort of moral perspective one intends, but this strategy of debasing the specificity of words has diminishing, and eventually negative, returns. “Rape” is only one example of this kind of black magic. When people do not realize this is what is occurring, their first strategy is almost always to double down, suggesting that whoever resists their redefining of terms is simply stupid, ignorant, and/or evil. And so initially, when resistance is encountered to a redefinition of terms, simply saying the same thing with more force is the strategy enjoined. This will work a few times until others realize this is what is going on, and fail to be impressed or cowed by that strategy. You can only use the word “racist” to mean “generally all-around despicable person” so many times, but once that term is used to the point of cliché, it fails to have the same power.
Resistance to black magic requires overcoming the superstitious association of the word. In just the way a pigeon will continue the behavior it has accidentally associated with a desirable outcome, many people are stupid and don’t realize that the power a word used to hold has been lessened. They have been trained to think that an appeal to democracy is a crowd pleaser, so once they encounter someone truly anti-democratic, they don’t know how to make head nor tail of their rejection, because to them “democracy” always just meant the positive affirmation it yielded from others; they never really thought it through for its denotative content. (Think “Freedom, justice, and the American way.” What the hell does that even mean?)
The “early adopters” of denotative substantialism (let’s call this the position that “you shouldn’t separate a word’s exosemantic and connotative content from its denotative content”) will almost always encounter this superstitious resistance. In fact, many who would take up this kind of substantialism never do, because they are afraid of encountering resistance from the mass of people who find the connoted and exosemantic too useful to give up its use to someone who only wishes to insist on maintaining its denotative content. In this way “racist” has lost essentially all denotative meaning (survey your friends about what it means, and see how compatible those meanings are with all the ways it is actually used or what else it means for those engaging in merely empirical, positive scientific analysis), without anyone noticing because anybody who would care about the actual denotative meaning of racist is probably a racist himself anyway.
In much the same way the alleviating of Malthusian pressure sees the proliferation of individuals who are of lower than median fitness among the population (and of which a disproportionately small amount will survive the next meeting with the Malthusian ceiling), words stand under the same pressure. If there are no selective pressures on the use of words, maladapted uses of words proliferate and contribute to a breakdown of communication. Given the difficulty of measuring the effectiveness of communication (like most things, we don’t notice something is getting worse until it completely breaks) it follows that small defections from the purpose of communication, rather than the politically opportune use of a word to attack opponents, can be used for a short time for political benefit. There is a limit to this, and eventually a sufficient amount of breakdown in the agreed use of words will lead to large scale defections and breaking of alliances which will see groups of people cease to be able to communicate effectively with one another their disagreements.